E D I T O R I A L C O M M E N TA R Y

Severe Community-Acquired Pneumonia (CAP) and the Infectious Diseases Society of America/American Thoracic Society CAP Guidelines Prediction Rule: Validated or Not
Lionel A. Mandell
Division of Infectious Diseases, Henderson Hospital, McMaster University, Hamilton, Ontario, Canada

Downloaded from http://cid.oxfordjournals.org/ at HINARI Peru Administrative Account on August 2, 2013

(See the article by Liapikou et al. on pages 37785)

Identifying patients with severe community-acquired pneumonia (CAP) who require admission to an intensive care unit (ICU) can, at times, be a difcult and daunting task. It is not always clear which patients will benet from the additional diagnostic, treatment, and management protocols and procedures of the ICU, and the consequences of a poor selection process can be disastrous. ICU facilities, resources, and personnel are relatively limited in most hospitals. Therefore, the inappropriate admission to the ICU of patients with CAP who do not require such care may prevent a patient who does require such care from accessing it. The subsequent transfer of patients with CAP who are rst admitted to a hospital ward to the ICU for delayed onset of respiratory failure or septic shock is associated with increased mortality [1]. To anyone who cares for patients who may have severe CAP, it is obvious that the course of the disease is dynamic and that
Received 21 October 2008; accepted 25 October 2008; electronically published 13 January 2009. Reprints or correspondence: Dr. Lionel A. Mandell, McMaster University/Henderson Hospital, Div. of Infectious Diseases, 711 Concession St., Fifth Fl., Wing 40, Rm. 503, Hamilton, Ontario L8V 1C3, Canada (lmandell@mcmaster.ca). Clinical Infectious Diseases 2009; 48:3868 2009 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2009/4804-0002$15.00 DOI: 10.1086/596308

neither clinical nor laboratory values remain static. It can be difcult to differentiate between individuals who require ICU care at the time of assessment in the emergency department and those whose conditions will worsen after admission to the hospital. Ideally, we would like to identify patients who require ICU care as early as possible. Having an accurate prediction rule that allows physicians to select patients with severe CAP who require ICU treatment early in the course of illness facilitates the appropriate initial management and antibiotic treatment and is an important strategy for mortality reduction [2]. The decision regarding site of care (i.e., whether the patient should be treated as an outpatient, in a hospital ward, or in the ICU) carries with it a number of important implications. It is for these reasons that having an accurate and reliable prediction rule is important. The site of care determines the type and extent of diagnostic testing, the spectrum and route of administration of antibiotics, and the overall treatment costs. Rules that are overly sensitive or insufciently specic help no one. A number of criteria have been developed over the years to help with the definition of severe CAP and/or to identify patients who require admission to an ICU.

These include the original American Thoracic Society (ATS) guidelines published in 1993 and the revised version published in 2001; the confusion, elevated blood urea nitrogen, respiratory rate, and blood pressure [CURB] score; the CURB plus age 65 years [CURB 65] score; and the Pneumonia Severity Index (PSI). All of these guidelines and measures attempted to deal with the concept of CAP severity [37]. Some, such as the CURB and CURB 65 scores, were in fact severity-ofillness scores, whereas the PSI was a prognostic model that was originally developed to identify patients who could be managed at home. Part of the problem has been that there has not been a universally agreed upon denition of severe CAP. An examination of North American guidelines published over the past 14 years shows a process that has been slowly but progressively evolving. The original ATS CAP guidelines listed 9 criteria, and the presence of any 1 of these criteria implied that the patient had severe CAP. Such an approach, however, resulted in a denition that was extremely sensitive but not specic [8]. The ATS guidelines of 2001 modied the denition of severe CAP to include the presence of 2 minor criteria (respiratory rate 30 breaths per min, ratio of arterial oxygen tension to inspired oxygen fraction !250, bilateral or

386 CID 2009:48 (15 February) EDITORIAL COMMENTARY

multilobar pneumonia, systolic blood pressure 90 mm Hg, and diastolic blood pressure 60 mm Hg) or the presence of 1 major criterion (the need for mechanical ventilation, septic shock or the need for vasopressors for 14 h, an increase in the size of inltrates by 150% within 48 h, and acute renal failure). A study by Angus et al. [9] compared the outcomes of hospitalized patients with CAP who received ICU care with the outcomes of those who did not. Angus et al. [9] compared the predictive characteristics of the original and revised ATS criteria, the British Thoracic Society criteria, and the PSI criteria for ICU admission, receipt of mechanical ventilation, medical complications, and death. They found that, with ICU admission and receipt of mechanical ventilation as the outcome measures, the revised ATS guidelines were the best predictor; when medical complications and death were the outcome measures, the PSI was the best predictor. It is important to note that the authors stipulate that, in both situations, none of the prediction rules were found to be particularly effective. Angus et al. [9] concluded that none of the available prediction rules for severe CAP were adequately robust to guide clinical care at the current time [9, p. 717]. Ewig et al. [10], in a subsequent article, conrmed the ability of the modied ATS rule to predict severe pneumonia. The Infectious Disease Society of America (IDSA)/ATS CAP guidelines are quite explicit about what constitutes major criteria for either severe CAP or direct admission to the ICU [11]. Either the need for mechanical ventilation with endotracheal intubation or the presence of septic shock requiring receipt of vasopressors are absolute indications. The minor criteria, however, are less clear-cut. A total of 9 such criteria are given in the guidelines, and the presence of 3 criteria was considered to provide sufcient evidence for admission to an ICU or high-level monitoring unit. The 9 criteria are respiratory rate 30

breaths per min, ratio of arterial oxygen tension to inspired oxygen fraction 250, multilobar inltrates, confusion and/or disorientation, uremia (blood urea nitrogen level 20 mg/dL), leukopenia (WBC count !4000 cells/mm3), thrombocytopenia (platelet count !100,000 platelets/ mm3), hypothermia (core temperature !36C), and hypotension requiring aggressive uid resuscitation. The reader is referred to the IDSA/ATS CAP guidelines for a discussion of the minor criteria and the reasons for their inclusion [11]. The study by Liapikou et al. [12] in this issue of Clinical Infectious Diseases is an attempt to validate the predictive rule suggested by the IDSA/ATS CAP guidelines for the identication of patients with severe CAP and the selection of those individuals who require ICU admission. The study is an important one from both academic and clinical standpoints, and it is the rst study, to our knowledge, to validate the recent prediction rule. The authors prospectively observed consecutive patients with CAP who met predened criteria. The study took place over a 7year period from January 2000 through January 2007, at which time the new guidelines were rst published online, followed shortly thereafter by publication in print. The IDSA/ATS prediction rule was retrospectively applied to the patient database, but such an approach should have no bearing on the results. The main outcomes of interest were the predictive capacity of severe CAP criteria for ICU admission and hospital mortality and the impact of ICU admission on hospital mortality for patients who met only minor severity criteria and no major criteria. For the relationship between severe CAP criteria and ICU admission, the sensitivity and specicity were 71% and 88%, respectively, whereas for mortality, the sensitivity and specicity were 58% and 88%, respectively. The rule tended to overestimate ICU admission somewhat, but overall, when compared with the modied ATS criteria of 2001, the IDSA/ATS prediction rule was equally good at predicting

ICU admission and better at predicting hospital mortality. As might be expected, severity determined on the basis of a major criterion had the strongest association with mortality. As for the predictive value of the minor criteria only, the authors were unable to document a reduction in mortality among patients who were admitted to the ICU, nor did the number of minor criteria present predict any benet from ICU admission. The authors concluded that the need for ICU management was clear when either of the major criteria were employed but that the need for ICU care when only the minor criteria were used was not unequivocally supported by their data. If we examine the IDSA/ATS criteria for severe CAP, the value of the major criteria is self evident. It goes without saying that a patient who requires intubation and mechanical ventilation or a patient with septic shock who requires vasopressors would need treatment in an ICU. The minor criteria, however, are not as obvious in terms of their relationship to mortality or the necessity for ICU care. I would agree with the authors when they state that the need for ICU admission derived from minor criteria alone is uncertain in our population and deserves further prospective evaluation [12, p. 377]. This is virtually identical to a statement made in the IDSA/ ATS guidelines themselves; when referring to the minor criteria, the committee wrote that prospective validation of this set of criteria is clearly needed [11, p. 539]. There are 2 questions that can be asked of the article by Liapikou et al. [12] that relate to the minor criteria and to 1 of the major criteria. We are told that 235 patients were admitted to the ICU and that this included 41 patients from other wards who were admitted to the ICU after their condition deteriorated. In the absence of any major criteria, how many and/or what types of the minor criteria did these specic 41 patients meet? It is also reported that 57 (43%) of the patients with septic shock were initially treated and stabilized in the emergency de-

Downloaded from http://cid.oxfordjournals.org/ at HINARI Peru Administrative Account on August 2, 2013

EDITORIAL COMMENTARY CID 2009:48 (15 February) 387

partment and did not require subsequent admission to the ICU. This seems like a high percentage of such patients to do so well. We are then told, however, that the poorer outcome in such patients conrms the need for close monitoring and ICU care of these patients [12, p. 383]. This suggests that too many patients with septic shock were admitted to hospital wards when they might have benetted from ICU admission instead. It is unfortunate that studies of ICU admission do not account for patients who have a do not resuscitate status. Such patients may, in fact, meet severity criteria and die without being considered for ICU admission. The article by Liapikou et al. [12] describes a nicely performed study that validates the IDSA/ATS prediction rule when it comes to major criteria but fails to conrm the validity of the minor criteria. These ndings are welcome but are not very surprising, and it is incumbent upon investigators to continue to explore the usefulness of the minor criteria.
Acknowledgments
Potential conicts of interest. L.A.M.: no conicts.

References
1. Leroy O, Santre C, Beuscart C, et al. A veyear old study of severe community-acquired pneumonia with emphasis on prognosis in patients admitted to an intensive care unit. Intensive Care Med 1995; 21:2431. 2. Kollef MH, Sherman G, Ward S, Fraser VJ. Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest 1999; 115: 46274. 3. Niederman MS, Bass JB, Campbell GD, et al. Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. American Thoracic Society. Medical Section of the American Lung Association. Am Rev Respir Dis 1993; 148: 141826. 4. Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001; 163:173054. 5. British Thoracic Society Research Committee. Community-acquired pneumonia in adults in British hospitals in 19821983: a survey of aetiology, mortality, prognostic factors, and outcome. Q J Med 1987; 62:195220. 6. Lim WS, van der Eerden MM, Laing R, et al. Dening community-acquired pneumonia severity on presentation to hospital: an international derivation and validation study. Thorax 2003; 58:37782. 7. Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with 8.

9.

10.

11.

12.

community-acquired pneumonia. N Engl J Med 1997; 336:24350. Ewig S, Ruiz M, Mensa J, et al. Severe community-acquired pneumonia: assessment of severity criteria. Am J Respir Crit Care Med 1998; 158:11028. Angus DC, Marrie TJ, Obrosky DS, et al. Severe community-acquired pneumonia: use of intensive care services and evaluation of American and British Thoracic Society Diagnostic Criteria. Am J Respir Crit Care Med 2002; 166:71723. Ewig S, de Roux A, Bauer T, et al. Validation of predictive rules and indices of severity for community-acquired pneumonia. Thorax 2004; 59:4217. Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/ American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007;44:S2772. Liapikou A, Ferrer M, Polverino E, et al. Severe community-acquired pneumonia: validation of the Infectious Diseases Society of America/ American Thoracic Society Guidelines to predict an intensive care unit admission. Clin Infect Dis 2009; 48:37785 (in this issue).

Downloaded from http://cid.oxfordjournals.org/ at HINARI Peru Administrative Account on August 2, 2013

388 CID 2009:48 (15 February) EDITORIAL COMMENTARY